Communicating an unclassifiable DNA variant — the hardest result in genetic counseling to deliver without being misread as a diagnosis or dismissed as nothing
A Variant of Uncertain Significance (VUS) is a real DNA difference from the reference genome that clinical laboratories cannot yet confidently classify as disease-causing or harmless. It is neither reassuring nor actionable — a genomic finding suspended in the middle of a five-tier confidence scale. Because patients rarely have prior exposure to genuinely inconclusive test results, a VUS is chronically misread in one of two opposite and equally dangerous directions: as a de facto diagnosis ("I have the gene for this disease") or as clinically meaningless noise ("this is nothing, forget about it").
Every variant identified by clinical sequencing is sorted into one of five ACMG/AMP categories: Benign, Likely Benign, Uncertain Significance, Likely Pathogenic, or Pathogenic. A VUS sits in that literal middle category — the laboratory has evidence, but not enough of it, and not consistent enough in direction, to confidently push the variant toward either pole.
Critically, a VUS is not a fourth kind of disease state. It carries no established clinical action: it should not trigger prophylactic surgery, should not by itself justify a change in screening intervals beyond what personal or family history already indicates, and should not be used as the sole basis for reproductive or major medical decisions. It is, in the most literal sense, an open scientific question about one specific DNA position — not yet an answer about the patient.
A rule of thumb used in genetic counseling training: a VUS should change almost nothing about clinical management on the day it is disclosed. If a clinical action is being taken purely because of a VUS, that is usually a sign the result is being over-interpreted.
The proportion of tests returning at least one VUS ranges from roughly 5% for small, well-studied single-gene tests up to 30% or more for large multi-gene hereditary cancer or cardiac panels, and rises further with exome or genome-scale sequencing. Two forces drive this:
• Panel breadth: every additional gene tested adds more positions where a rare, poorly characterized variant can be found — more genes simply means more chances to land on uncertainty. • Ancestry and reference bias: population reference databases (gnomAD and predecessors) are still disproportionately built from individuals of European ancestry. A variant that is common and clearly benign in an underrepresented population may appear "rare" in these databases purely due to under-sampling, artificially inflating VUS rates for patients from those populations. This is a recognized, actively-being-corrected equity gap in clinical genomics.
Because "uncertain" is an unfamiliar concept in most patients' experience of medical testing (where results are usually "positive" or "negative"), a VUS tends to be forced into one of two more familiar — and both incorrect — mental models:
• Over-interpretation: the patient hears "variant" and "gene associated with disease X" and concludes they have or will get disease X, sometimes pursuing unnecessary surgery, screening, or profound anxiety. • Under-interpretation: the patient hears "uncertain" and "we don't know," and concludes the entire test was inconclusive or worthless, disengaging from any follow-up — including the reclassification monitoring that could later become clinically meaningful.
Both failure modes trace back to the same root cause: a VUS result is being asked to behave like a binary diagnostic answer when its actual epistemic status is a probability distribution that the field is still refining.
A VUS designation is not a shrug — it is the output of a structured, points-based evidence framework. The ACMG/AMP guidelines (Richards et al., 2015, and subsequent ClinGen refinements) define specific categories of evidence, each independently assessed and weighted, that are combined using defined rules to arrive at one of the five classification tiers. Understanding these categories is what lets a counselor explain, concretely, why a variant is stuck in the middle rather than leaving the impression that "nobody has looked at this."
Laboratories assemble evidence for a variant from four broad streams, each contributing independent — and sometimes conflicting — signal:
• Population frequency: how common is this exact variant in large reference cohorts (e.g. gnomAD, spanning hundreds of thousands of sequenced individuals)? A variant present at an allele frequency higher than plausible for the disease's prevalence argues toward benign; true absence from large healthy-population databases is weak-to-moderate evidence toward pathogenic, never proof.
• Computational / in silico prediction: algorithms (REVEL, CADD, SIFT, PolyPhen-2, AlphaMissense, and others) estimate whether an amino acid change is likely to disrupt protein structure or function based on evolutionary conservation and biophysical modeling. These are probabilistic predictions, not measurements — they earn only "supporting" strength evidence on their own.
• Functional / experimental studies: cell-based or biochemical assays that directly test whether the altered protein behaves abnormally (e.g. loss of enzymatic activity, mislocalization, failure to bind a partner protein). This is some of the strongest evidence available, but validated functional assays exist for only a minority of genes and variants.
• Segregation and case-control data: does the variant travel together with the disease through multiple generations of an affected family (segregation), and is it enriched in disease cohorts versus controls? This requires large, well-documented pedigrees or case series that frequently do not yet exist for a rare novel variant.
A variant becomes Likely Pathogenic or Likely Benign only when the combined weight of evidence, following defined combining rules, clears a specific threshold in one direction. A VUS is the default outcome whenever:
• Evidence exists on both sides and roughly cancels out (e.g. reassuring computational prediction but concerning absence from population databases) • Evidence exists in only one direction but is too weak in isolation to meet the threshold (e.g. only a single supporting-strength computational prediction, nothing else) • No meaningful evidence exists yet at all — a genuinely novel variant with no case reports, no functional data, and population frequency data too sparse to interpret
This is the key message for disclosure: a VUS does not mean the evidence points to danger and was ignored. It usually means the evidence is either balanced, thin, or simply not yet collected — and that more data, not a different opinion, is what would resolve it.
ClinGen Variant Curation Expert Panels now formally re-weight and standardize these evidence codes gene-by-gene, because a criterion calibrated generically across all genes can be miscalibrated for any single one — this specialization is part of why reclassification rates keep improving over time.
How a VUS is delivered shapes patient understanding far more than the underlying molecular biology does. Genetic counseling literature consistently shows that framing, pacing, and repetition of the "uncertain — not a diagnosis" message determine whether a patient walks away appropriately informed or walks away either needlessly frightened or falsely reassured. Disclosure is a skill, not a formality.
Best practice favors a layered structure rather than leading with molecular jargon:
1. Orient first: state plainly that a variant was found, and that it falls into an "uncertain" category — before any gene name or technical detail is introduced. 2. Define uncertainty explicitly: "This is different from a positive or negative result. It means we cannot yet say whether this specific change in your DNA causes increased risk or is a harmless difference." 3. State what does NOT change: management, screening, and family recommendations stay based on personal and family history — not on this result — unless and until it is reclassified. 4. Offer the mechanism for future resolution: explain that laboratories periodically re-review VUS findings as new evidence accumulates, and that the patient will be recontacted if the classification changes. 5. Provide it in writing: a plain-language letter or after-visit summary reinforces the spoken conversation, since recall of a single verbal disclosure is unreliable, especially under any emotional load.
Contrast the layered approach against a hypothetical blunt delivery: "You have a variant in [gene]. We don't know what it means." Delivered quickly, without framing or repetition, this pattern reliably produces a measurable distress spike and — paradoxically — worse long-term understanding. Patients under acute distress narrow their attention onto the gene name and the disease it is associated with, and lose the uncertain qualifier entirely. The nuance that should have been the headline becomes the first thing forgotten.
The counseling literature frames this as an asymmetry: the cost of a few extra minutes spent contextualizing "uncertain" is small; the cost of a patient who leaves believing they have a genetic diagnosis they do not have — or who disengages entirely from a result that may matter later — is large and difficult to correct after the fact.
A useful test before ending a disclosure conversation: ask the patient to restate, in their own words, what the result means. If they describe it as either "I have the disease" or "it's nothing," the framing has not landed yet, and it is worth revisiting before the visit ends.
Avoid: "mutation" (carries connotations of damage; prefer the neutral "variant" or "genetic change"), "abnormal gene," or any phrasing that implies a completed diagnosis. Avoid minimizing language too — "probably nothing" overstates certainty in the benign direction just as much as alarmist language overstates it toward pathogenic.
Prefer: "a change in your DNA that we can't yet classify," "not currently considered a cause of your condition," "this may be updated as research continues," and explicit numeric or qualitative framing of what is and is not yet known. Consistency in terminology across the verbal conversation, the written report, and any follow-up materials meaningfully reduces confusion.
When a pathogenic variant is found, testing at-risk relatives ("cascade testing") is a cornerstone of hereditary risk management. A VUS breaks that pattern: because the variant's clinical meaning is not yet established, a relative's test result — positive or negative for that same variant — cannot be reliably interpreted either, and testing them offers little to no actionable information while carrying real potential for confusion and anxiety.
Cascade testing works for a classified pathogenic variant because a relative's result directly answers a clinically actionable question: do they carry the risk allele or not? For a VUS, that same question has no actionable answer either way — a relative found to carry the same VUS still does not have a known increased risk, and one who does not carry it has not been "cleared" of anything the family didn't already know from history.
Outside of a formal, IRB-approved segregation study — where testing multiple family members is specifically how additional evidence gets generated to eventually help classify the variant — professional guidelines (including ACMG and NSGC positions) generally recommend against routine clinical cascade testing of relatives for a VUS. The potential harms — family anxiety, insurance and psychological questions, and a false sense that "we now know who is at risk" — outweigh the clinical benefit, because there usually is no clinical benefit yet to weigh against them.
The distinction to communicate clearly to families: "we are not testing your relatives for this specific change because a result either way would not tell us anything useful yet — not because we think it is unimportant."
Rather than testing relatives, the recommended path is twofold:
• Manage relatives' risk based on family and personal history alone, exactly as would be done if the sequencing test had never been performed — the VUS does not add or subtract from that risk assessment until it is reclassified. • Keep the family's contact information current with the ordering clinic or laboratory, since a future reclassification of the proband's variant may eventually make family communication and, at that point, targeted cascade testing appropriate and useful.
In selected situations — most often within research studies specifically designed to gather segregation evidence — relatives may be invited to test under a research protocol. This is conceptually different from clinical cascade testing: the relative's result there is not primarily for their own risk assessment, but a data point contributing to eventually classifying the variant for everyone.
A VUS is not a permanent label — it is a snapshot of the evidence available at one point in time. As population databases grow, new functional studies are published, and more families with the same or similar variants are identified worldwide, laboratories periodically re-evaluate stored VUS findings. Most are eventually reclassified toward Benign; a minority move toward Pathogenic. Either direction triggers a formal amended report and a recontact obligation.
Empirically, when VUS findings are tracked over years and eventually reclassified, the majority end up downgraded to Likely Benign or Benign rather than upgraded to Likely Pathogenic or Pathogenic. This reflects the underlying biology of the search: labs cast a wide net across thousands of positions in dozens of genes, and most rare variants encountered are, in fact, tolerated variation rather than disease-causing changes — they simply hadn't yet accumulated enough population or functional data to prove it at the time of the original report.
This is itself useful context to share during disclosure: while no individual VUS can be assumed reassuring in advance, the aggregate historical pattern leans toward eventual reassurance, not toward eventual bad news — this can temper the anxiety generated by "uncertain" without overstating what is known about any one specific variant.
Reclassification is driven by an accumulating evidence pipeline: laboratories and ClinGen Variant Curation Expert Panels continuously ingest newly published case reports, growing population reference databases, updated computational predictors, and, where available, new functional assay results. When enough new evidence crosses the ACMG/AMP classification thresholds, the variant's tier is formally updated in the laboratory's database and in public resources such as ClinVar.
Re-review is not always on a fixed calendar — some laboratories run continuous automated monitoring against updated databases, others perform batch re-analysis on a periodic (often one-to-two-year) cycle, and some rely on the ordering clinician or patient to request updated re-analysis. This variability is itself a known gap in the field, and part of why patients should be told explicitly how (and whether) they will be proactively recontacted, rather than assuming it will happen automatically.
When reclassification occurs, the laboratory issues a formal amended report superseding the original — and best practice is a proactive recontact of the patient (directly or via the ordering clinician) to schedule a renewed counseling conversation, not simply a passive database update the patient is expected to discover on their own.
If reclassified to Likely Benign/Benign: prior VUS-driven caution can generally be relaxed; management reverts fully to personal and family history-based guidelines.
If reclassified to Likely Pathogenic/Pathogenic: the finding becomes clinically actionable — management guidelines specific to the associated condition now apply, and, at this point, cascade testing of at-risk relatives becomes appropriate and is typically offered.
In the interim — however many years that takes — patients are counseled to: keep contact information current with the testing laboratory or clinic, mention the VUS at future genetics visits so re-analysis can be explicitly requested if it has not happened automatically, and continue any screening already indicated by personal and family history, independent of the variant's uncertain status.